I want to look at the clash in perceptions about race between forensic science and biological anthropology. We need to start off by talking about two dead white men: Carl Linnaeus (1707 – 1778) and Charles Darwin (1809 – 1882). Both of them wrote about human races, and their ideas continue to affect us in the modern world.

Linnaeus trained as a medical doctor, but his importance was in setting up the binomial system of classification (Homo sapiens was one of Linnaeus’ terms).  He imagined that every living organism had a specific type that was fixed; it was a once off thing that didn’t change.  His method of categorising  types into clusters of related species is known as typology.

In Linnaeus’ view every person came from a pure racial type that could be measured in anatomical terms. But Charles Darwin grasped something that Linnaeus didn’t: the idea of variation. Darwin realised that everybody varied a little bit, and that types were not clear and neat.

This concept of Darwinian variation on one side and Linnaean fixity on the other is why we are confused about our vision of race.

Should we perceive race as the Linnaean type with each kind of person around the world clearly defined, and with all people within each race looking very similar to one another? Or should we see them in the Darwinian model which grasps variation, where there are dynamic, ever-changing populations based on genetic difference?

A debate was held in 2007 in Albuquerque in the USA where biological anthropologists argued their view of human variation. While the majority of biological anthropologists were trying to deal with the nature of variation, the forensic scientists were saying you can look at a skull and see whether it’s a Mongoloid, a Negroid, or a Caucasoid person – very clear, neat racial categories.

The majority of scientists agreed that it was impossible to reconcile clear, separate identities of races in the light of genetic knowledge. As you move from place to place, you don’t cross a line somewhere in Western India, for instance, and find all people on one side are Mongoloid and everyone on the other is Caucasoid. It doesn’t work that way. There is a gradation.

But some scientists at the debate argued that there is regional variation which broadly parallels the racial categories.  For instance, people in China do look different to people in, say, the Sudan.  An American colleague of mine calls this the “any fool can see” theorem. Any fool can see that people are not the same in different parts of the world.  So who is right?  Are there distinct racial categories or are there not?

In forensics – and amongst the lay public – there’s a tendency to put people into discrete categories. Yes there are physical differences around the world, but individual variation makes it nearly impossible to use the physical differences to put people into neat groups. 

The traditional racial categorisation used in forensics includes three clear racial classifications based on average physical features found in Europe, Asia and Africa. These are regional patterns in human morphology with differences in shape that loosely reflect the genetic history of people.

The problem with it is that you can’t demonstrate the clarity of those definitions if you measure the skulls because of a gradation of difference. You can’t draw on a map a clear line separating groups of people.  The use of the racial categories instead forces you to ignore the variation and find a “best fit” for your unknown skull.

This means that the racial categories are partly in the mind of the observer, but from the police perspective, these categories are real. Yet they’re stereotypes, as if everyone from Asia looks like a separate Linnaean type from everyone in Europe or in Africa. Variation simply isn’t taken into account. The race definitions as used in forensics don’t have a solid scientific basis in genetics, they’re perceived variation rather than real and measureable.

There are also problems with sliding definitions. The physical anthropologists are talking about biological features, but racial categories incorporate cultural features: the headscarf worn by a Muslim woman, or hair straightening practices. Police aren’t talking specifically about biology, they’re talking about bureaucratic and political definitions of races which incorporate non-biological aspects

A “popular” belief or classification of race is called “folk taxonomy”, what people believe makes up a race. This folk classification is a mixture of biology and culture – technically it’s about ethnicity versus race, culture versus biology.

When it comes to forensic cases, the issue of racial identification is further confused by the so-called “CSI effect” (after the television show Crime Scene Investigation). The public is being trained in forensic science and forensic anthropology by the popular media.

Juries and victims of crime now have the idea that forensic science works just like it does on CSI: the crime can be solved by tomorrow afternoon at 2pm. Science and scientists are seen as infallible, giving the public an unrealistic expectation of the science.

Scientific racial identification is such a contentious issue that the public perception of its meaning is heavily influenced by the CSI effect.

Take facial reconstruction of an unidentified person, using clay to build up face over a skull. This gives the impression that we can put a face back on a skull. Actually it’s facial approximation. About two thirds of the facial reconstruction is science, based on the underlying bone structure; one third is art, where variations like ear lobes, skin texture, skin colour and facial fat are added by the artist and you don’t know if it’s reflective of the original face.

The CSI effect convinces the public that the scientists are capable of replicating the original face as an exact reconstruction, yet the objective was never to get a perfect reflection of what that person looked like.  The real purpose was to trigger a response where someone from the public comes forward with another line of evidence which will lead to identification of that person. It’s a last-chance method when everything you’ve used to identify that missing person hasn’t worked.

The CSI effect has also had an impact in the arena of DNA. There’s this idea about a gene that can identify the race of a person. It is simply not true. Genes tell us about a person’s ancestor, not that person’s race. For example, you might track the mitochondrial DNA, the maternal lineage, to find that a person’s ancestor was African. The mitochondrial DNA lineage ignores all of the other genetic history of the person and gives no information about appearance.  All it says is that the person’s maternal lineage goes back to Africa. This is useful for the anthropologist but it’s not useful for the forensic specialist.

These conflicts occur because you’re dealing with an academic subject that is applied in the real world. The problem isn’t whether or not people differ; it’s how we perceive that categorisation that we use to label people. Science doesn’t always have the same meaning out there on the street as it does in the lab. It can be difficult to bridge the two. Forensic science needs unambiguous information for use in court cases but with race, how scientists perceive the range of variation in people around the world is very different from the way the lay public and the police see it. There’s a clash between what you expect to see and what science can demonstrate.

Professor Alan Morris of the Department of Human Biology at the University of Cape Town, who recently launched his book Missing & Murdered: a Personal Adventure in Forensic Anthropology, spoke at the Africa Genome Education Institute’s final lecture of 2011 in the Darwin Seminars series.